Support AUTO_INCREMENT for MySQL and AUTOINCREMENT for SQLite

[miuy56dc]

In MySQL it's AUTO_INCREMENT
(see https://dev.mysql.com/doc/refman/8.0/en/create-table.html)
and in SQLite it's AUTOINCREMENT.

We use ColumnOption::DialectSpecific(Vec<Token>) to avoid adding a new variant for each vendor-specific column option.

[coveralls]

Pull Request Test Coverage Report for Build 186220604

• 37 of 38 (97.37%) changed or added relevant lines in 5 files are covered.
• 204 unchanged lines in 3 files lost coverage.
• Overall coverage increased (+0.2%) to 91.827%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
src/parser.rs	1	2	50.0%

Files with Coverage Reduction	New Missed Lines	%
tests/sqlparser_common.rs	25	96.81%
src/ast/mod.rs	47	81.09%
src/parser.rs	132	88.91%

Totals
Change from base Build 175941714:	0.2%
Covered Lines:	4393
Relevant Lines:	4784

---

💛 - Coveralls

[maxcountryman]

This looks good to me. 👍

@nickolay does this seem okay to you?

[nickolay]

Yes, looks good to me too, but please add a note about this being MySQL-specific as a comment in the code (better in all of: AST, the parser and the tests), as:

• PostgreSQL has SERIAL
• MSSQL has IDENTITY [ (seed, increment) ]
• ANSI uses GENERATED { ALWAYS | BY DEFAULT } AS IDENTITY [ ( ... ) ]

Documenting the differences between dialects like this should be useful both for our users and for designing a better solution for the dialects problem.

AUTO_INCREMENT also seems to be a good case for trying ColumnOption::DialectSpecific(Vec<Token>), as suggested earlier, what do you think?

[miuy56dc]

Thank you for your suggestions. I find SQLite support AUTOINCREMENT, so add MySQLAutoIncremnt and SQLiteAutoIncrement to solve the problem.

[nickolay]

Thanks!